FeynmanBench: leaderboard
Metric: Score (%), the mean of the five checkpoint pass rates (vertex inventory, propagator inventory, topological connectivity, momentum routing, full amplitude) on FeynmanBench's Feynman-diagram tasks (a rendered diagram plus minimal conventions; Standard Model QED, electroweak and QCD diagrams from tree level to two loops), API models at temperature 0, mean of five runs; four models ran on subsets of 165 to 742 items, the others on about 2,000; higher is better. Source: arxiv.org. Saturation forecast: Not forecast. 19 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 44.3 |
| 2 | Gemini 3 Flash | 40.2 |
| 3 | Qwen 3.6 Plus | 29.6 |
| 4 | Seed 2.0 Pro | 28.6 |
| 5 | Claude Sonnet 4.6 | 28.5 |
| 6 | Qwen 3.5 Plus | 27.7 |
| 7 | GLM-5V Turbo | 21.3 |
| 8 | GPT-5.1 | 12.4 |
| 9 | GPT-5.4 | 12 |
| 10 | Claude Opus 4.7 | 12 |
| 11 | GPT-5.2 | 11.1 |
| 12 | Claude Opus 4.6 | 10.7 |
| 13 | Claude Opus 4.5 | 10.5 |
| 14 | Kimi K2.5 | 3.2 |
| 15 | Claude Sonnet 4.5 | 2 |
Interactive version: theaggregate.ai/benchmark?slug=feynmanbench · How It Works · Data refreshed daily, snapshot 2026-10-07.